Bug #12512
closed
Netgate Hardware (SG-1100 - SG3100) preloaded firmware issues
Added by Tim Brogdon almost 4 years ago.
Updated almost 4 years ago.
Category:
Operating System
Affected Architecture:
SG-1100, SG-2100, SG-3100
Description
I have noticed on all of our Netgate hardware we have to re-install the OS upon receiving, failure to do so prevents access to check for updates or install packages. Further if we continue to run eventually the firewall will corrupt completely, causing downtime and a full reload to get it working again. I have opened a few support tickets and the first recommendation is to to reload the OS due to corruption. After the reload all issues appear to be resolved.
I suspect the image being used to load the SG line is corrupt/faulty. I have confirmed this on at least 5 units, possibly more as we did not know what the issue was initially.
A tech mentioned this can happen if no WAN is connected, that probably should be a bug report of it's own, but I can confirm with a SG-1100 I setup today, the first thing I did was connect both WAN and LAN before applying power and the same issue happened directly after going through the wizard. My usage is extremely basic on these units for small offices. There are no vlans or special configuration.
- Status changed from New to Closed
Tim,
Unfortunately we can't replicate anything like that update check issue here, and there isn't enough detail to classify exactly what you are seeing as a general bug. There appear to be at least two separate and likely unrelated issues in your description. It's possible some of that is due to filesystem corruption, but that would not typically affect the update or package checking process in the way you describe above.
That said, there are some known issues with UFS which tend to result in similar problems with corruption, such as when the units lose power abruptly, and we are working on making ZFS the default filesystem for most devices which will effectively work around that type of problem. That's already being tracked separately, so this would be a duplicate of that issue, and unlikely to be a problem on future releases.
As for the update check, if you can reproduce the problem again open an issue with TAC and have them gather some diagnostic information, such as a status output, and dig into it further. Their fist suggestion to reload is in the interest of getting customers up and running again ASAP, but depending on the symptoms there may be more we can do to narrow down what might be happening if it doesn't turn out to be related to the filesystem. Since we can't replicate that kind of issue here, we would need more detail from a unit in the problematic state to diagnose the problem.
Also available in: Atom
PDF
Updated by